Given their history of family adversity, foster children are a high-risk population for negative sibling relationships and frequent physical aggression toward a sibling leading to increased risk for physical aggression toward other children i.e., non sibling peers. The foster placement context provides a critical window for timely and amenable youth violence prevention to alter the downward behavioral trajectories of children with histories of neglect. This study uses the context of the sibling relationship to develop and evaluate the effectiveness of a youth violence prevention intervention. The goal of this application is to reduce sibling aggression which is prevalent in neglected children placed in foster homes and forecasts significant risk for youth violence. Sibling aggression among maltreated children remains a highly under identified and untreated problem. The failure to intervene represents a major public health concern. This intervention utilizes a tran-stheoretical model that implicates child vulnerabilities (altered cortisol and emotion dysregulation), sibling pair risk- (low positively, and poor conflict management), and foster parent- (non mediation) factors into a testable intervention model to decrease risk for sibling and peer physical aggression among young neglected children. The 'Promoting Sibling Bonds' intervention combines two promising short-term psychosocial sibling programs: Sibs Together (Bank, 2000;2004) and Mediation-Based Intervention Strategy for Sibling Conflict (Smith &Ross, 2007) adapted to meet the needs of neglected children in the foster care system. The study involves two phases. In Phase 1 Intervention Development, we will refine and expand the intervention protocol, adapt and integrate manuals, train therapists (interventionists), make changes from iteration work, and try-out behavioral outcome measures. In Phase 2 Implementation of a controlled trial, we will recruit a sample of N = 144 children (nested in 72 pairs) between ages 5-8 years old;collect baseline data (wave1);randomize sibling pairs and their families (foster parent or biological parent for cases who reunite) to intervention and comparison groups, and collect post-intervention (wave2) and 6 months follow-up (wave3) outcome data of the adapted intervention. We hypothesize that at post-intervention (wave2), and again at follow-up (wave3), as compared to children assigned to the comparison 'usual care '(UC), children assigned to the experimental intervention condition (IC) will show improvement in reported sibling aggression, non sibling peer aggression, emotion dysregulation;and children will be more likely to exhibit a typical diurnal cortisol pattern. Sibling pairs will show improved relationship quality (positivity and negativity). The proportion of foster parents and sibling pairs who engage in conflict mediation, compromise, and reconciliation strategies will be higher in the intervention group. The public health impact of this intervention for violence prevention is high. Our innovative outcome model is a part of a second generation of studies aimed at integrating rich behavioral science and advances in neurobiology to better understand the multilevel change mechanisms in prevention trials.